π A Complete Team of AI Analysts for Trading#
TradingAgents simulates the dynamics of a real trading firm β but with LLM agents instead of humans. π€
π’ The Framework Structure#
Just like a real investment firm:
Analyst Team:
- π Fundamentals Analyst β evaluates company financials
- οΏ½οΏ½ Sentiment Analyst β aggregates news, StockTwits, and Reddit
- π° News Analyst β interprets macro event impacts
- π Technical Analyst β detects patterns with MACD, RSI, etc.
Research Team:
- π Bullish and bearish researchers in structured debate
Management:
- π Trader Agent β makes decisions based on all reports
- βοΈ Risk Management β evaluates volatility and liquidity
- πΌ Portfolio Manager β approves/rejects transactions
π οΈ Multi-LLM and Open Source#
Supports: OpenAI, Anthropic, Google, xAI, DeepSeek, Qwen, GLM, and local Ollama.
π‘ Explanation in a nutshell#
TradingAgents is an open-source multi-agent framework for financial trading research with LLMs. It replicates the structure of a real investment firm: specialized analysts (fundamental, sentiment, technical), researchers in debate, a trader agent, and risk management. Agents debate each other to find the optimal strategy. Built with LangGraph, it supports dozens of LLM providers and includes a decision log with continuous learning. β οΈ Not financial advice.
More information at the link π
